From aeb50dff605b648fca341ebae728287ee082ffbc Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Marcel=20M=C3=A4rtens?= Date: Wed, 25 Nov 2020 21:42:33 +0100 Subject: [PATCH] Drop is run only AFTER each statement. And the guard generated here in the if let Some() will live through the WHOLE statement, which is definitly critical. The fix is quite easy. We just move it out in a seperate line. There are still some participants where the dropping will fail, and followup disconnects will NOT get triggered and be queued up till ethernity, but at least we are not blocking new logins --- network/src/scheduler.rs | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/network/src/scheduler.rs b/network/src/scheduler.rs index ab648c9acc..33cb4ed054 100644 --- a/network/src/scheduler.rs +++ b/network/src/scheduler.rs @@ -270,7 +270,9 @@ impl Scheduler { // 3. Participant will try to access the BParticipant senders and receivers with // their next api action, it will fail and be closed then. trace!(?pid, "Got request to close participant"); - if let Some(mut pi) = self.participants.lock().await.remove(&pid) { + let pi = self.participants.lock().await.remove(&pid); + trace!(?pid, "dropped participants lock"); + if let Some(mut pi) = pi { let (finished_sender, finished_receiver) = oneshot::channel(); pi.s2b_shutdown_bparticipant_s .take() @@ -278,8 +280,9 @@ impl Scheduler { .send(finished_sender) .unwrap(); drop(pi); - trace!("dropped bparticipant, waiting for finish"); + trace!(?pid, "dropped bparticipant, waiting for finish"); let e = finished_receiver.await.unwrap(); + trace!(?pid, "waiting completed"); return_once_successful_shutdown.send(e).unwrap(); } else { debug!(?pid, "Looks like participant is already dropped");